$50,000 Computer!?

Rohan K -

Hello everyone! I finished writing my random search algorithm to find the best parameters for the Voice Activity Detector. However… on the iMac that I’m working on, to search 1,000 parameter combinations it would take an estimated 330,000 HOURS! That’s 37 years.

Thankfully, the lab has a $50,000 Lambda Vector computer that can run my program faster. After lunch, I will test how long it takes to run one search. Based on that, I will choose a number of searches that will have time to run overnight. The goal is to have a new set of parameters tomorrow that produces a lower error than the original parameters. I’m very curious to see the difference in processing speed between a $1,000 iMac and a $50,000 computer. If it can save me 37 years, I guess it’s worth it.

While I wait for the script to run, I wanted to give a brief overview of what it’s like to work in the lab at Mass Eye and Ear. The whole experience is priceless, and I’ve learned so much from venturing into the real world. The lab is a part of Mass General Brigham, the third best hospital in the world, and through Harvard Medical School the lab has several spaces on the fourth floor of Mass Eye and Ear. When I come in to work at 9am, I can see doctors and patients in the lobby. On the fourth floor, my computer faces two windows overlooking the Charles River. I start the day eating breakfast while I review the code that I wrote the day before. Then there’s hours of troubleshooting and brainstorming code that takes into account all special cases in the dataset. Aki, the post-doc from Finland sits 20 feet away, so I ask him any questions I have. Having someone who has a PhD in computer science has helped me learn how to code in just two weeks. At noon, the whole team eats lunch on the seventh floor, where we discuss the most random topics. Most of the researchers are international, so they share interesting stories and new perspectives. By 1pm we are back to work. Occasionally, we’ll have lab meetings where we discuss the projects we’re working on and ask any questions. By 5:30pm, I wrap up any code that needs to be run over night, and then I take the train home. I’ll give an update next week on how the random search algorithm went and how long it took on the really expensive computer.

More Posts

Comments:

All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.

    akhil_d
    That's really interesting Rohan! I am glad to see you are learning a lot and enjoying your time in Boston. How can the Lambda Vector computer run your program faster? Also, do you know how common it is for other labs to have the Lambda Vector?
    March 18, 2024 at 4:27 pm -
    rohan_k
    Hi Akhil! The Lambda Vector runs much faster because it has a 32 core processor vs the iMac which has a 4 core processor. For labs that deal with machine learning and high volume programs, a large capacity computer like Lambda is essential.
    March 19, 2024 at 7:32 am -
    Bryn Sharp
    Rohan, Do you think there is a way to make faster processing speeds more accessible to researchers who do not have access to a lab with a $50,000 computer? Or, do you think there is a way to re-program your program so that it does not require such fast processing speeds? Just food for thought. I'm super excited to hear about the progress you are making. Keep up the great work!
    March 20, 2024 at 11:20 am -

Leave a Reply